Loading ToolBench
An MCP server with OAuth 2.0 authentication provider running on Cloudflare Workers, demonstrating remote MCP capabilities with authorization workflows
This server defines a single tool 'add' with Zod schema validation. The tool has a basic description and two numeric parameters with minimal descriptions. While the schema is present and typed, the descriptions are too brief to meet LLM-optimized standards, and critical documentation patterns are absent. The tool lacks error handling guidance, output schema documentation, and contextual information about when or why to use it. No patterns from the 54 Agentic Tool Patterns are explicitly followed.
Adds two numbers together
Tool description is too brief (14 characters). LLM-optimized descriptions should be 50-200 chars, explaining WHAT the tool does, WHEN to use it, and what it returns.
Parameter descriptions are minimal (12 and 17 characters respectively). Each parameter needs 50+ character descriptions explaining constraints, format, and purpose.
Output schema is not documented. The tool description does not specify what the response contains, forcing LLMs to infer the structure. Document that it returns a text response with the sum.
No error handling guidance. If invalid inputs are passed (e.g., non-numeric values), there is no recovery instruction for the LLM. Add error classification and recovery hints.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | F | 46 | 2026-07-28+ | v2 |
| 2026-03-09 | F | 29 | - | v1 |
Tool name 'add' lacks action-verb clarity. While 'add' is a verb, it does not follow the convention 'verb_noun' (e.g., 'add_numbers'). For a math operation, the name should be more specific.
No numeric constraints documented. Parameters 'a' and 'b' are numbers, but there is no description of valid ranges, precision, or handling of special values (negative, zero, very large numbers).